AI Disclosure and Responsible Use Notice

What the system does

The model predicts responses from your Input. It does not understand facts or intent as a human does, does not independently verify its answer, and does not act on your behalf unless your own application gives it tools or permissions. Model configuration and infrastructure may change; the current model and price are displayed in the Service.

Material limitations

• Outputs may be false, incomplete, outdated, biased, offensive, unsafe, or inconsistent.
• Generated code may contain security vulnerabilities or destructive behavior. Review and test it in an isolated environment before execution.
• Outputs may resemble third-party material. You are responsible for checking rights, attribution, and permitted use.
• The model is not a substitute for a qualified professional and does not provide verified legal, medical, financial, tax, safety, or other regulated advice.
• Do not use an Output as the sole basis for a decision that affects a person's legal rights or access to employment, credit, insurance, housing, education, healthcare, or essential services.

Human oversight and user responsibility

Use a competent human reviewer, verify important claims with authoritative sources, apply domain-specific safeguards, and provide clear AI disclosure to people affected by your implementation where appropriate or legally required. You are responsible for the prompts, tools, data, permissions, deployment context, and actions of applications you build with the Service.

Data handling

Netra does not use your inputs or outputs for training, fine-tuning, advertising, its own content-based evaluations, or model or service improvement. This applies to covered API requests, temporary Playground and Free OCR, including prompts, code, files, tool calls/results and responses.

Content recording is off by default. Optional saved content is deleted within 30 days of original capture/Save, including recoverable backups; editing or viewing does not reset the deadline. Native shared RAM/GPU cache entries can remain for seven days from creation without reset. Per-workspace opt-out blocks shared-cache reads and new writes once activated; existing entries remain until their original expiry.

The Inference Data Handling Policy explains Netra's use of "ZDR", caching and opt-out limits. The Privacy Policy explains account data, operational metadata, support access and fallback processing.

Safety and reporting

Do not submit children's data, credentials, or sensitive personal information unless you have a lawful and appropriate basis. Do not attempt to bypass safeguards or use the Service for prohibited conduct described in the Terms of Service.

Report a harmful output, security issue, or suspected misuse to rifky@netraruntime.com and include the request ID when available. Do not include secrets or unnecessary personal data in the report.